Pancreatic multipotent progenitors (MPCs) that give rise to all islet endocrine cells, including insulin-producing beta cells, emerge within the early pancreatic bud epithelium. During this time, the pancreatic epithelium undergoes a dramatic and transient stratification, with epithelial cells losing their polarity as they generate a multilayered structure, but then quickly regaining it as the epithelium resolves and pancreatic branching begins [4]. The mechanisms and consequences of this rapid morphological change, and of the underlying dynamic shifts in cell polarity, are completely unknown. In support, current thinking in the field has shifted towards acknowledging the potential impact of 3D architecture on beta cell fate [5]. The hypothesis driving this work proposes that the process of rapid stratification and de- stratification of the pancreatic epithelium during development, along with the associated changes in cell polarity, directly impacts endocrine cell fate. This proposal aims to characterize cell polarity during pancreatic stratification and resolution, and to test whether the polarity determinants Numb, Numb-like and Par3 are required for pancreatic morphogenesis, MPC specification and endocrine differentiation. Understanding the stepwise processes by which endocrine beta cells acquire their fate and differentiate into functionally mature insulin-producing cells will advance efforts towards cell replacement therapies to treat diabetes, as these steps will be mimicked for in vitro differentiation or regeneration.